Sunita Tarai

I've built a real-time chat application on the MERN stack, enabling instant messaging, user login, chat rooms, and notifications through WebSocket communication.

I also developed RetinaAI, an AI-powered retinal disease detection system that classifies OCT scans into CNV, DME, Drusen, and Normal categories using MobileNetV3 transfer learning. The model was trained on 84,495 retinal images and achieved 97.33% training accuracy and 95.3% validation accuracy.

For RetinaAI, I created a Flask REST API for real-time inference, SQLite3 patient-history logging, automated PDF medical reports, and a vanilla JavaScript single-page interface with drag-and-drop uploads and color-coded severity triage.

I'm an MCA graduate with foundations in Java, Python, full-stack web development, database management, and machine learning. I'm seeking a challenging role where I can apply my technical skills, problem-solving ability, and interest in innovation while continuing to grow.